The relationship between yield and each of its attributes and some physiological traits in grain
sorghum under well-watered and drought stress conditions
Mohammed A. Sayed
Agronomy Dept., Faculty of Agriculture, Assiut University, Egypt
Corresponding author: [email protected]
Abstract
It is desirable for sorghum breeder to know the extent of relationship between yield and each of
morphological and physiological traits affecting it which facilitate breeder in selecting plants of
desirable traits, especially under drought stress conditions. Therefore, this work was conducted at two
locations represent clay and sandy soils of the Experimental Farms of Faculty of Agriculture, Assiut
University, Assiut, Egypt, during two summer seasons 2014 and 2015 under two water regimes. To
obtain the well-watered conditions (WW), surface irrigation was applied in the clay soil and
genotypes were watered each 14 days, while drip irrigation system was used in the sandy soil and
plants were watered for 2 hours each 3 days. To induce drought stress conditions (DS), the third and
the fifth surface irrigations were skipped in the clay soil, while plants were drip irrigated for 1 hour
each 3 days in sandy soil. Three statistical procedures i.e., simple correlation coefficient, the path
coefficient analysis and the stepwise regression analysis were performed to determine the functional
relationships between yield and each of its attributes and some physiological traits under both
treatments. For this, 43 grain sorghum genotypes including 30 F1 hybrids, their elven parents (6
female and 5 male lines were crossed in line x tester mating fashion) and two check cultivars were
used in this study. Results revealed highly significant differences among genotypes for all studied
traits under WW and DS conditions. Panicle weight had the highest positive correlation with grain
yield of sorghum genotypes under both treatments followed by threshing percentage and seed index,
reflecting these traits are the most contributed to yield. Path analysis showed that panicle weight and
threshing percentage had positive and direct effect on grain yield/plant, while chlorophyll content,
excised leaf water loss and stay green were the most important physiological traits under DS
conditions. In addition, panicle width and chlorophyll content showed the highest positive indirect
effects on grain yield. Stepwise regression exhibited that panicle weight and threshing percentage had
the strongest variation in grain yield per plant under both water regimes. On the other hand, all
physiological traits except excised leaf water loss (under WW conditions) showed positive correlation
coefficient with GY/P under both treatments. Stepwise regression revealed that relative water content
was the most important physiological trait followed by flag leaf area under WW conditions, while
chlorophyll content was the most important physiological trait under DS followed by excised leaf
water loss that contributed high amount of the total variation of grain yield.
Keywords: Correlation, Path coefficient, Stepwise regressions, Drought stress, Sorghum
1
Introduction
Globally, drought stress is a serious environmental stress that affects more frequently plant growth
and productivity due to the current climate change scenario (Fracasso et al 2016). Among world’s
major cereal crops, Sorghum (Sorghum bicolor L. Moench) is the fifth important one and is valued
for its grain, stalks and leaves. It is cultivated in Egypt as a summer crop and is concentrated in the
middle and upper parts (Hassanein et al 2010). In upper Egypt and in semi-arid regions as well,
sorghum is often exposed to drought and heat stress, which affect plant growth and grain yield (Prasad
et al 2008). Sorghum crop shows considerable variation in agronomic, morphological and
physiological traits that response to selection and are highly influenced by environmental factors
(Ezeaku et al, 1997). Sorghum improvement to drought tolerance and the other stresses requires from
breeders to concentrate on utilization of desirable traits that may aid in superior improved cultivars
aiming to surpass the present productivity level (Warkad et al 2010). Efforts are currently focused to
increase the cultivated area of sorghum in Upper Egypt in newly reclaimed desert land (Ali 2012) by
growing high yielding-drought tolerant varieties.
In cereal crops, yield depends on a combination of morphological and physiological attributes that
affect the yield directly or indirectly way under well-watered and drought stress conditions. Among
these, plant height, panicle length, panicle width, panicle weight, threshing percentage, grain numbers
and 1000-grain weight (Warkad et al 2010; Ghasemi et al 2012; Tolk et al 2013; Khaled et al 2014).
In addition, excised-leaf water loss (Wang and Clarke, 1993, Ahmad et al 2009); relative water
content (Rad et al 2013, Ahmad et al 2009); stay green (Subudhi et al 2000, Borrell et al 2014);
chlorophyll content (Brito et al 2011, Asadi et al 2015); flag leaf area (Asadi et al 2015) as examples
of the physiological traits and have been measured to assess drought tolerance in sorghum.
Raising grain yield potential is one of the major objectives in sorghum improvement programmes that
can be achieved via improving of yield components and physiological traits (Golparvar 2013).
However, yield is a quantitative and very complex trait that results of the interaction between various
yield attributes, this interaction varies according to the environment which the plant lives. Several
statistical approaches, such as correlation, path coefficient and stepwise regression analysis are very
helpful and beneficial in explaining the relationship between yield and contributing factors, especially
under different environmental conditions. Further, these tools provide the ability to study the
interrelationships and inter-dependence among traits. In multivariate analysis, the plant attributes are
referred to as the independent variables, while yield is the dependent variable. Each of the
independent variables contributes to the variation in the yield of the genotypes (Abd El-Mohsen
2013). Nature and the magnitude of correlation coefficients of the traits help breeders to determine
the selection desirable criteria for simultaneous improvement of various traits along with yield.
2
However, selections based on simple correlation coefficients without considering the interactions
among yield and yield attributes may mislead the breeder to reach his main breeding purposes (Del
Moral et al, 2003). Therefore, path coefficient analysis that was derived from Wright (1921) and
described by Dewey and Lu (1959) provides an aid for partitioning of correlation coefficient into
direct and indirect effects of different traits on yield and thus helps in assessing the cause-effect
relationship as well as effective selection. Also, stepwise multiple linear regression aims to construct
a regression equation that includes the independent variables accounting for the majority of the total
yield variation. Determination of the relationship between yield and its attributes was reported in
several published researches (for instance, Ezeaku and Mohammed 2006; Jain et al 2010; Chavan et
al 2011 and Abubakar and Bubuche 2013). However, physiological screening of sorghum germplasm
for drought response is limited (Rakshit et al 2016). Yield and a number of physiological traits have
been used to select drought tolerant genotypes (White et al 1994). Detailed measurements of the yield
components and the physiological traits under a range of water stress conditions are needed to obtain
and better understand the possible combination of these traits as independent variables to enhance
yield under drought conditions. Therefore, the objective of this study was to increase the
understanding of relationships between grain sorghum yield and each of morphological and
physiological traits under well-watered and drought stress conditions by studying correlations,
stepwise multiple regression and path analysis.
Material and Methods
Experimental sites and plant materials
This investigation was carried out at two locations of the experimental farms of the Faculty of
Agriculture, Assiut University, at Assiut, Egypt, during the two summer successive seasons, 2014
and 2015 under two water regimes. The first location was at Faculty of Agricultural Research Farm,
Assiut University (AS location), while the second location was at the newly reclaimed area at the
Experimental Station of the Faculty of Agriculture, Al-Wadi Al-Assyouti Farm (WAD location),
Assiut University (25 km South East of Assiut). Details of the physical and chemical properties of
the two locations were described in Sayed and Mahdy (2016) and Sayed and Bedawy (2016).
30 F1 grain sorghum crosses formed by crossing six inbred lines (cytoplasmic male sterility lines) to
five testers in a line × tester mating design in the summer season of 2013 in addition two standard
checks (Hybrid 305 and Dorado) for comparison were used in this study. The female lines ICSA.11,
ICSA.329, ICSA536, ICSA598, ICSA625 and ATXA629 and male lines ICSR102, ICSR59,
ICSR628, ICSR89013 and ICSR 89034 were obtained from India (International Crop Research
Institute for Semi-Arid Tropics, ICRISAT) except one female line (ATXA629) was obtained from
Texas A&M University, College Station, TX, U.S.A.
3
Experimental design and water regimes
Two separate field treatments (well-watered and drought stress treatments) were performed at each
location. The experimental design was a strip plot design in a randomized complete block
arrangement with three replications. Water regimes were allocated to the main plots and genotypes
to subplots. Each genotype was placed in a one row plot of 3 m long and 0.6 m apart with 0.2 m
between plants. Trial was hand planted with 3-4 seeds per hill, which was later thinned to two plants
per hill. Planting was done in the two summer successive seasons at the 17th and 18th of June, 2014
and in 16th and 17th of June, 2015, in the first and second locations, respectively. Standard cultural
practices for optimum sorghum production were carried out at each location. In the first location, to
obtain well-watered conditions (WW), entries were watered using surface irrigation each 14 days and
as recommended for optimum sorghum production, while to obtain drought stress conditions (DS),
the third and the fifth irrigations were skipped. In the second location, in both treatments drip
irrigation was used and plants were watered each 3 days. In WW treatments, plants were irrigated for
2 hours while in DS treatment, plants were irrigated for 1 hour (drought stress conditions started after
30 days from sowing and continued until fully ripening).
Recording of observations
Grain yield/plant (GY/P; g) and its attributes were recorded on five tagged guarded plants from each
plot. The yield attributes were; plant height (PH; cm), panicle length (PL; cm), panicle width (PW;
cm), panicle weight (PWG; g), threshing percentage (TH %) and Seed index (SI; g) whereas days to
50% blooming (DB) data were recorded on whole plot basis. In addition, five physiological traits
related to drought tolerance were measured in this study, namely relative water content (Barrs 1968);
excised-leaf water loss (Clarke 1987); chlorophyll content (Xu et al 2000); stay green (Wanous et al
1991) and Flag leaf area (FLA) (Montgomery, (1911). Details of the measurements of the yield, yield
attributes and the physiological traits of were described in (Sayed and Mahdy 2016 and Sayed and
Bedawy 2016).
Statistical analysis and procedures
The combined analysis of variance as outlined by Gomez and Gomez (1984) was computed for wellwatered and drought stress treatments separately after carrying out the homogeneity of variances
using Bartlett test using SAS software (v 9.2, 2008). Phenotypic correlations among yield attributes
and the physiological traits along with grain yield were determined under well-watered and drought
stress conditions separately across years and locations using Pearson’s correlation test using SAS
software. Path coefficient analysis was performed under the two water regimes and between yield
and each of its attributes and the physiological traits separately using Analysis of Moment Structures
software (AMOS v. 5; Arbuckle 2005). The direct and indirect effects of influential variables on grain
yield were calculated according to proposed method of Dewey and Lu (1959). Stepwise linear
4
regression according to Draper and Smith (1966) was computed using SAS software to determine the
appropriate variables significantly contributed to total variation in yield and the relative contribution
was calculated as (R2). In path coefficient and stepwise regression analyses, grain yield was examined
as the dependent variable versus other traits as independent variables.
Results and Discussion
Analysis of variance of the studied traits under both treatments
Data in Tables (1 and 2) show the combined analysis of variance and some summary statistics for
grain yield, its attributes and some physiological traits related to drought tolerance of 43 sorghum
genotypes under well-watered and drought stress conditions, respectively across two locations and
over two seasons. ANOVA revealed highly significant differences between seasons, between both
locations and for their interaction for the majority of the investigated traits, reflecting the impact of
the environmental conditions on the expression of the investigated traits of the genotypes. Likewise,
results showed highly significant differences among genotypes for all studied traits under wellwatered and drought stress conditions, indicating the existence of sufficient variability among
genotypes. In addition, the interaction of genotypes with years was significant for all studied traits
under both treatments, while the interaction of genotypes with locations and the triple interaction
were insignificant for all studied traits except view cases under both treatments. From the combined
analysis, all of investigated traits showed wide range of variability under both treatments. For instants,
GY/P ranged between 21.1 and 59.7 g under well-watered conditions and between 13.8 and 53.1 g
under drought stress conditions. ELWL as an example for the physiological traits ranged from 49 to
79.9% under well-watered conditions and from 49.6 to 85.4% under drought stress conditions.
Therefore, the presence of such range of variations of the traits indicated that the presence of large
amount of genetic variation among tested lines, hybrids and check cultivars, which is the source of
variable genetic material. The coefficient of variability (C.V.) of the investigated traits (Tables 1 and
2) was higher under drought stress conditions than under well-watered conditions, may due to the
different responses of the genotypes to drought stress. Since, days to 50% heading exhibited the
minimum percentage of coefficient of variation (2.6%) under both treatments, while flag leaf area
showed the maximum percentage of coefficient of variation (27 to 29.1%) under well-watered and
drought stress conditions, respectively. Tariq et al (2007) also reported higher phenotypic variance
for grain yield among the sorghum varieties. Also, Tag El-Din et al (2012) and El-Naim et al (2012)
reported that highly significant differences were obtained among grain sorghum genotypes for yield
and its attributes. Amare et al (2015) studied the variability for yield, yield related traits and
association among traits of sorghum varieties in Ethiopia and found a wide range of variation among
5
the varieties across locations for yield and its attributes this variation confirmed by high values of
phenotypic and genotypic variation for the investigated traits.
Correlation among studied traits under both treatments
Data in Table (3) shows the correlation coefficients among the studied traits under well-watered and
drought stress conditions over years and locations. Under well-watered conditions, the analysis
revealed that grain yield/ plant (GY/P) was associated positively and significantly (P ≤ 0.01) with all
studied traits except ELWL the correlation (r=-0.07), which was negative and insignificant. Among
yield attributes, the strongest correlation coefficient with GY/P was obtained by panicle weight
(0.94**) followed by threshing percentage (r=0.66**), indicating that these traits are the most
contributed to yield/plant. There was positive and significant correlation between seed index and each
of plant height (r=0.51**), Panicle length (r=0.31**), panicle width (r=0.14**), panicle weight
(r=0.19**) and threshing percentage (r=0.30**). Relative water content (RWC) had the highest
correlation coefficient but moderate (r=0.36**) with GY/P among the physiological traits followed
by flag leaf area (r=0.27**), reflecting the significance of water maintain in leaves tissues for yield
production. It has been observed that RWC was correlated positively and highly significantly with
the yield attributes traits, e.g. TH% (r=0.50**) and SI (r=0.27**). Meanwhile, RWC correlated
significantly and positively with each of chlorophyll content (r=0.26**), stay green (r=0.28**) and
flag leaf area (r=0.29**). However, no evidence of a relationship between RWC and ELWL under
well-watered conditions. Chlorophyll content was associated positively and significantly with each
of plant height, panicle traits, relative water content and flag leaf area. Under drought stress
conditions, the same trend was observed for the association between grain yield and each of its
attributes and the physiological traits with an exception that the coefficients were much higher under
drought stress than under well-watered conditions. RWC was correlated positively and significantly
with GY/P and its attributes except days to 50% heading, the correlation was negative. Also, positive
and significant correlation between RWC and CC was observed (r=0.34**). Tag El-Din et al (2012)
found positive correlation coefficients between grain yield per plant and each of panicle length,
panicle width, 1000-kernel weight and leaf area. Khaled et al (2014) reported associations between
grain yield per plant and its attributes in sorghum. Amare et al (2015) found that GY/P showed high
positive and significant correlation with panicle weight per plant, leaf area index, plant height and
1000-seed weight. Also, Ezeaku and Mohammed (2006) reported high positive phenotypic
correlation coefficients of grain yield with head weight and plant height across two locations. Our
findings are in partial agreement with the previous mentioned reviews, therefore, the positive
correlation of grain yield per plant with studied traits suggested that the possibility of simultaneous
improvement of grain yield per plant through indirect selection of these positively correlated traits.
6
Stepwise multiple regression
Yield attributes under both treatments
In stepwise regression analysis, grain yield was examined as the dependent variable versus its
attributes as independent variables under well-watered and drought stress conditions. The hierarchical
stepwise regression involved four steps (models), Table (4) shows the outcome of this regression. In
the first step, panicle weight was the most important trait and had the strongest variation in grain yield
per plant and accounted 87.5% of the variance in GY/P. The second step, threshing percentage entered
the next after PWG and accounted 9.84% of the variance. The third step, days to 50% heading came
the third predictor and accounted 0.16% of the variance in GY/P. The final step, seed index entered
the last and accounted 0.03% of the variance, the final model could justify significantly more than
97.5% changes in performance of GY/P (R²=87.5%) according to the equation:
GY/P; g = -39.42+0.57 PWG + 0.55 TH% + 0.13 50% DH + 0.06 SI
Under drought conditions, the stepwise regression analysis involved three steps. In these steps, three
variables namely panicle weight, threshing percentage and days to 50% heading from the previous
analysis under well-watered were remained in the final model and seed index was excluded. Panicle
weight accounted 91.10% of the variance in GY/P followed by threshing percentage which explained
6.54% and finally days to 50% heading accounted 0.19% of the variance. The independent variables
in the final model justified 97.83% of the GY/P variance. Seed index of sorghum grains is affecting
by the starch accumulation during flowering and filling stages and drought stress causes reduction in
the total accumulation of starch in the grains (Emes et al 2003 and Bing et al 2013). Saed-Moucheshi
et al (2013) and Nasri et al (2014) reported that spike weight per unit had a positive and significant
regression coefficient on grain yield in wheat. Seed index was affected negatively by drought stress
and that may be a reason for discarding seed index from the final model. Therefore, the best prediction
equation was formulated as follows:
GY/P; g = -32.08+0.62 PWG + 0.41 TH% + 0.14 50% DH
Physiological traits under both treatments
Table (5) shows the accepted physiological traits as independent variables and their relative
contributions in relation of grain yield/plant variance under well-watered and drought stress
conditions. Since, the stepwise regression analysis revealed two models under well-watered and three
models under drought conditions. The results showed that relative water content (RWC) and flag leaf
area (FLA) R2 = 15.5%, had justified the maximum of yield changes under well-watered conditions.
RWC was the most important physiological trait followed by FLA, hence, the relative contributions
in the total variation of grain yield were 13.02% and 2.13%, respectively. The low relative
contributions of the physiological traits may due to these traits are not the components of sorghum
7
yield in fact, but have significant correlation with grain yield. RWC and FLA appeared to be important
traits for good production under well-watered conditions. Consequently, based on the final step of
stepwise regression analyses, the best prediction equation was formulated as follows:
GY/P; g = -4.68+0.53 RWC + 0.04 FLA
Under drought conditions, the scenario of the physiological traits was changed because of excluding
RWC and FLA from the final step of the stepwise regression analysis and adding other physiological
traits namely; chlorophyll content (CC), excised leave water loss (ELWL) and stay green (Stg).
However, these traits had justified the maximum of yield changes (R2 =32.88%). Chlorophyll content
was the most important physiological trait under drought conditions and explained 29.29% of the
variance in GY/P followed by excised leaf water loss that contributed 2.67% of the total variation of
grain yield, and finally stay green was accounted small effect (0.92%) of the total variance of GY/P.
Consequently, based on the final step of stepwise regression analyses, the best prediction equation
was formulated as follows:
Grain yield function equation GY/P; g=-53.49+0.19 ELWL+1.41 CC+ 1.31 Stg
Saed-Moucheshi et al (2013) found that chlorophyll content of the flag leaf and leaf area had a
positive and significant regression coefficient on grain yield in wheat. Tolk et al (2013) found that
panicle mass and leaf area were important traits under drought stress in sorghum.
Yield attributes and the physiological traits together under both treatments
Table (6) shows the final step resulted from stepwise regression analysis between yield as dependent
variable and the all studied traits as independent variables under both treatments. It can be seen that
seven traits out of twelve were involved in the final model of stepwise regression analysis under wellwatered conditions. The model accounted about 97.78% of the variance and panicle weight and
threshing percentage explained 87.50 and 9.80% of the variance in GY/P, respectively. Excised leaf
water loss explained about 0.30% of the variance in GY/P among physiological traits. Therefore, the
best prediction equation was formulated as follows:
GY/P; g=-34,44+0.58 PWG + 0,55 TH% + 0,05 ELWL-0,07 CC -0,05 FLA +0,05 HD + 0,047 SI
Under drought stress conditions, the stepwise analysis showed that the final step included six traits
out of twelve, these traits were the most contributed to GY/P with discarding seed index from the
final model. Since, panicle weight explained 91.10% of the variance in GY/P followed by threshing
percentage which accounted 6.50% of the variance in GY/P. Chlorophyll content explained about
0.20% of the variance in GY/P among physiological traits. Panicle mass at maturity provides an
integration of growth conditions between the flag leaf stage and the start of grain filling, which is
8
considered to be a part of the critical period for seed number determination (van Oosterom and
Hammer 2008) and consequently grain yield. Therefore, the best prediction equation was formulated
as follows:
GY/P; g=-35,79+0.61 PWG+ 0,38 TH% + 0,13 HD + 0,09 CC -0,005 FLA + 0,018 ELWL
Path coefficients analysis
For yield attributes
The estimates of direct and indirect effects of the seven yield attributes and five physiological traits
on grain yield/plant under well-watered and drought stress conditions are presented in Tables 6 and
7. Path coefficient analysis was performed using coefficient of all the traits with grain yield
plant/plant. Results revealed that panicle weight, threshing percentage, days to 50% heading and seed
index exerted positive direct effect on grain yield (0.810**, 0.310**, 0.040** and 0.020**,
respectively) under well-watered conditions (Table 7). The highest indirect effects on grain yield were
observed with panicle width (0.388) followed by threshing percentage (0.324). In addition, panicle
weight (2.187) and threshing percentage (0.862) had the highest total effects on grain yield /plant
under well-watered conditions. While under drought conditions (Table 7), the same trend was
observed, since panicle weight, threshing percentage and days to 50% heading showed positive direct
effect on grain yield/plant. Seed index had insignificant effect on grain yield compared to its effect
under well-watered conditions. Panicle width (0.569) and threshing percentage (0.424) showed the
highest indirect effects on grain yield/plant. Furthermore, panicle weight (0.963), threshing
percentage (0.726) and panicle width (0.723) showed the highest total effects on grain yield/plant
under drought stress conditions. Yield attributes accounted about 98% of the variance in grain
yield/plant under both treatments (Figure 1 A and B). on the other hand, the direct effect of the
residual was around 0.15 under both treatments. These findings are in partial agreement with those
obtained by (Arunkumar et al 2004, Premlatha et al 2006, Warkad et al 2010, Chavan et al 2011,
Abubakar and Bubuche 2013 and Khaled et al 2014) who found one or more of yield attributes affect
directly or indirectly on grain yield in sorghum.
For the physiological traits
Among studied physiological traits, relative water content and flag leaf area had the highest direct
effects on grain yield under well-watered conditions (Table 7 and Figure 2 A) and exerted 0.310**
and 0.150**, respectively. Both traits gave the highest total effects on grain yield/plant (0.568 and
0.216, respectively). Stay green had the highest indirect effect followed by flag leaf area on grain
yield/plant and gave 0.088 for each (Table 7 and Figure 2 A). While under drought stress, chlorophyll
9
content exhibited the highest positive direct effect (0.510**) on grain yield/plant followed by excised
leaf water loss (0.170**) then by stay green (0.100**). In addition, chlorophyll content, stay green
and relative water content gave 0.546, 0.245 and 0.180 as highest total effects on grain yield/plant
under drought conditions, respectively. It was observed that chlorophyll content (0.172) and stay
green (0.135) showed the highest positive indirect effect on grain yield/plant under drought conditions
(Table 7 and Figure 2 B). The physiological traits under study explained around 16 and 33% of the
variance in grain yield/plant under well-watered and drought stress conditions, respectively (Figure
2 A and B). On the other hand, the direct effect of the residual was high for the physiological traits
as independent variables under well-watered (0.92) and drought stress (0.82) conditions (Figure 2 A
and B), indicating the inadequacy of the trait chosen for the path analysis. Arunah et al (2015) stated
that leaf area index was the most contributed trait to sorghum yield as an importance photosynthetic
ability of a plant as an index of assimilates production for yield. Saed-Moucheshi et al (2013) reported
positive indirect effect of chlorophyll content and leaf area on on grain yield in wheat. Ali et al (2009)
found that flag leaf area exhibited strong positive relationship with grain yield and plays a vital role
in drought tolerance.
For yield attributes and the physiological traits together under both treatments
The direct and indirect effects of the yield attributes and the physiological traits to grain yield under
well-watered and drought stress conditions are presented in Table (9). Results revealed that panicle
weight, threshing percentage, days to 50% heading and excised leaf water loss exerted significant and
positive direct effect on grain yield (0.819**, 0.336**, 0.019** and 0.043**, respectively), while
panicle width, chlorophyll content and flag leaf area showed significant and negative direct effect on
grain yield/plant under well-watered conditions. The highest indirect effects on grain yield were
observed with panicle width (0.575) followed by threshing percentage (0.328). While under drought
conditions, panicle weight, threshing percentage, days to 50% heading and chlorophyll content
exerted significant and positive direct effect on grain yield (0.799**, 0.294**, 0.040** and 0.038**,
respectively), while panicle width and flag leaf area showed significant and negative direct effect on
grain yield/plant. The highest indirect effects on grain yield were observed with threshing percentage
(0.418) followed by chlorophyll content (0.382). These results are in partial agreement with those
obtained by Ali et al (2009); Tolk et al (2013) and Saed-Moucheshi et al (2013).
Grain yield of sorghum is the integration of various variables that affect plant growth throughout the
growing period. Many efforts have been achieved to develop proper models that can predict grain
yield and distinguish the ideal- and high yielding crop plants (Saed-Moucheshi et al 2013). The
knowledge of association and relationship between grain yield and its components under water stress
conditions would improve the efficiency of breeding programs by identifying appropriate indices to
10
select sorghum genotypes. The results of the present study showed highly significant differences
among genotypes for all studied traits under well-watered and drought stress conditions. However,
the existence of sufficient variability among genotypes may help sorghum breeders in investigating
and understanding the association between yield and the influencing variables under both water
regimes. Also, results showed that panicle weight had the highest positive correlation with grain yield
of sorghum genotypes under both treatments followed by threshing percentage and seed index,
reflecting these traits are the most contributed to yield. Panicle weight and threshing percentage had
the strongest variation in grain yield per plant and accounted 87.5 and 9.84% of the variance in GY/P
under well-watered conditions, respectively. While they explained 91.10 and 6.54% of the GY/P
variance under drought stress conditions, respectively. On other hand, all physiological traits except
excised leaf water loss (under well-watered conditions) showed positive correlation coefficient with
GY/P. This refers to that the genotypes that show high water status, high chlorophyll content and high
leaf area can be considered most tolerant to drought than others. But stepwise regression revealed that
relative water content was the most important physiological trait followed by flag leaf area under
well-watered conditions, while chlorophyll content was the most important physiological trait under
drought conditions followed by excised leaf water loss that that contributed high amount of the total
variation of grain yield. Selections based on simple correlation coefficients without considering the
interactions among yield and the independent attributes may mislead the breeder to reach his main
breeding purposes (Del Moral et al, 2003). Therefore, path coefficient and stepwise regression are
more informative than correlation because of separating the direct effects from the indirect effects
and determine the variables accounting for the majority of the total yield variation, respectively.
Conclusions
It can be concluded that among yield components, panicle weight and threshing percentage were
critical for maintaining yields under drought conditions. While chlorophyll content, excised leaf water
loss and stay green were the most important physiological traits under drought stress conditions. In
addition, these traits showed the highest direct positive effects on grain yield under drought stress,
while panicle width and chlorophyll content showed the highest positive indirect effects on grain
yield/plant. Therefore, selection can be done under drought stress conditions for these traits as
selection criteria for drought tolerance.
11
References
Abd El-Mohsen, A.A., 2013. Correlation and regression analysis in barley.Scientific Research and
Review Journal, 1(3), pp.88-100.
Abubakar, L. and T.S. Bubuche (2013) Correlation analysis of some agronomic traits for biomass
improvement in sorghum (Sorghum Bicolor L. Moench) genotypes in North-Western
Nigeria. African J. Agric. Res. 8(28), 3750-3756.
Ahmad, R.T., T.A. Malik, I.A. Khan and M.J. Jaskani (2009) Genetic analysis of some morphophysiological traits related to drought stress in cotton (Gossypium hirsutum). Int. J. Agric.
Biol, 11, 235-240.
Ali M.A. (2012) Effectiveness of selection in the F3 and F5 generations in grain sorghum. Asian
Journal of Crop science .4:23-31. doi:103923/ajcs20122331.
Ali, M.A., A. Abbas, S. Niaz, M. Zulkiffal and S. Ali (2009) Morpho-physiological criteria for
drought tolerance in sorghum (Sorghum Bicolor) at seedling and post-anthesis stages. Int. J.
Agric. Biol.,11: 674–680
Amare, K., H. Zeleke and G. Bultosa (2015) Variability for yield, yield related traits and association
among traits of sorghum (Sorghum Bicolor (L.) Moench) varieties in Wollo, Ethiopia. J. Plant
Breed. & Crop Sci. 7(5), 125-133.
Arbuckle, J. L. (2005) Amos (Version 5.0) [Computer Program]. Chicago: IBM SPSS.
Arunah, U.L., U. F. Chiezey L. Aliyu and A. Ahmed (2015) Correlation and Path Analysis between
Sorghum Yield to Growth and Yield Characters. Journal of Biology, Agriculture and Healthcare.
5 (19).
Arunkumar, B., Biradar, B. D. and Salimath (2004). Genetic variability and character association
studies in rabi sorghum. Karnataka J. agric. Sci. 17: 471-75.
Asadi, A.A., Valizadeh, M., Mohammadi, S.A. and Khodarahmi, M., (2015) Genetic analysis of
some physiological traits in wheat by generations mean analysis under normal and water deficit
conditions. In Biological Forum (Vol. 7, No. 2, p. 722).
Barrs, H.D., (1968) Determination of water deficits in plant tissues. In: T.T. Kozolvski (Ed.), Water
Deficits and Plant Growth, Vol 1, pp 235–368. Academic Press, New Delhi.
Bing, Y.I., ZHOU, Y.F., GAO, M.Y., ZHANG, Z., Yi, H.A.N., YANG, G.D., Wenjuan, X.U. and
HUANG, R.D., 2014. Effect of drought stress during flowering stage on starch accumulation
12
and starch synthesis enzymes in sorghum grains. Journal of Integrative Agriculture, 13(11),
pp.2399-2406.
Borrell, A.K., Mullet, J.E., George-Jaeggli, B., van Oosterom, E.J., Hammer, G.L., Klein, P.E.
and Jordan, D.R., (2014) Drought adaptation of stay-green sorghum is associated with canopy
development, leaf anatomy, root growth, and water uptake. Journal of Experimental Botany,
p.eru 232.
Brito, G.G.D., Sofiatti, V., Lima, M.M.D.A., Carvalho, L.P.D. and Silva Filho, J.L.D., (2011)
Physiological traits for drought phenotyping in cotton. ActaScientiarum. Agronomy, 33(1),
pp.117-125.
Chavan, S.K., Mahajan, R.C. and FATAK, S.U., 2011. Correlation and path analysis studies in
sorghum. Crop Research, 42(1), p.2.
Clarke, J.M., (1987). Use of physiological and morphological traits in breeding programmes to
improve drought resistance of cereals. In: Srivastava, J.P., E. Porceddu, E. Acevedo and S.
Varma (eds.), Drought Toleranace in Winter Cereals, pp: 171–190. John Wiley and Sons, New
York
Del Moral, L.F., Rharrabti, Y., Villegas, D. and Royo, C., 2003. Evaluation of grain yield and its
components in durum wheat under Mediterranean conditions. Agronomy Journal, 95(2), pp.266274.
Dewey DR, Lu KH (1959). A correlation and path coefficient analysis of components of crested
wheatgrass seed production. Agron. J. 51:515-518.
Draper N. R. and H. Smith (1966) Applied Regression Analysis, Wiley, New York, NY, USA
El Naim, A.M., Ibrahim, I.M., Rahman, M.E.A. and Ibrahim, E.A., 2012. Evaluation of some
local sorghum (Sorghum bicolor L. Moench) Genotypes in Rain-Fed. International Journal of
Plant Research, 2(1), pp.15-20.
Emes M J, Bowsher C G, Hedley C, Hedley C, Burrell M M, Scrase-Field E S E, Tetlow I J.
2003. Starch synthesis and carbon partitioning in developing endosperm. Journal of Experiental
Botany, 54, 569-575.
Ezeaku I.E., Mohammed S.G. (2006). Character association and path analysis in grain sorghum. Afr.
J. Biotechnol. 5(14): 1337-1340.
13
Ezeaku IE, Gupta SC, Prabhakar VR (1997). Classification of sorghum germplasm accessions
using multivariate methods. Afr. Crop Sci. J.7: 97-108.
Fracasso, A., Trindade, L. and Amaducci, S., 2016. Drought tolerance strategies highlighted by two
Sorghum bicolor races in a dry-down experiment. Journal of plant physiology, 190, pp.1-14.
Ghasemi, A., Ghasemi, M.M. and Pessarakli, M., 2012. Yield and yield components of various
grain sorghum cultivars grown in an arid region.Journal of Food, Agriculture &
Environment, 10(1), pp.455-458.
Golparvar R.A. (2013): Genetic control and combining ability of flag leaf area and relative water
content traits of bread wheat cultivars under drought stress condition. Genetika, Vol 45, No. 2,
351-360.
Gomez, K.A. and A.A. Gomez (1984). Statistical procedures for agricultural research. John Wiley &
Sons.
Hassanein M.S., A.G. Ahmed and N.M. Zaki (2010). Growth and productivity of some sorghum
cultivars under saline soil condition. Journal of Applied Sciences Research. 6(11): 1603-1611.
Jain, S.K., Elangovan, M. and Patel, N.V., 2010. Correlation and path coefficient analysis for
agronomical traits in forage sorghum [Sorghum bicolor (L.) Moench]. Indian Journal of Plant
Genetic Resources, 23(1), p.15.
Khaled, A. G. A.; Aml A. Tag El-Din and Eatemad M. Hussein (2014) Correlation, Path Analysis
and RAPD Markers in Sorghum (Sorghum bicolor L. Moench) Genotypes. Assiut J. Agric. Sci.,
(45) No. (4) 2014 (15-28)
Montgomery, E.G., (1911) Correlation studies in corn. NebraskaAgrExpStaAnnu Rep, 24: 108-159
Nasri R., A. Kashani, F. Paknejad, S. Vazan and M. Barary (2014) Correlation, path analysis and
stepwise regression in yield and yield component in wheat (Triticum aestivum l.) under the
temperate climate of Ilam province, Iran. Indian Journal of Fundamental and Applied Life
Sciences. 4: 188-198.
Prasad, P.V.V., Pisipati, S.R., Mutava, R.N. and Tuinstra, M.R. (2008) Sensitivity of grain
sorghum to high temperature stress during reproductive development. Crop Sci., 48: 1911–1917
Premlatha, N., Kumaravadivel, N. and Veerabadhiran, P. (2006). Correlation and path analysis
for yield and yield traits in sorghum [Sorghum bicolor (L.) Moench]. Res. on Crops 7 : 187-90.
14
Rad, M.R.N., Kadir, M.A., Yusop, M.R., Jaafar, H.Z. and Danaee, M., (2013) Gene action for
physiological parameters and use of relative water content (RWC) for selection of tolerant and
high yield genotypes in F2 population of wheat. Australian Journal of Crop Science, 7(3), p.407.
Rakshit, S., Swapna, M., Dalal, M., Sushma, G., Ganapathy, K.N., Dhandapani, A.,
Karthikeyan, M. and Talwar, H.S., 2016. Post-flowering drought stress response of post-rainy
sorghum genotypes. Indian Journal of Plant Physiology, 21(1), pp.8-14.
Saed-Moucheshi, A., Pessarakli, M. and Heidari, B., 2013. Comparing Relationships among Yield
and Its Related Traits in Mycorrhizal and Nonmycorrhizal Inoculated Wheat Cultivars under
Different
Water
Regimes
Using
Multivariate
Statistics. International
Journal
of
Agronomy, 2013.
SAS Institute (2008). The SAS System for Windows, release 9.2. Cary NC: SAS Institute
Sayed M.A. and I.M. Bedawy (2016) Heterosis and inheritance of some physiological criteria
imparting drought tolerance of grain sorghum in the irrigated and water-limited environments.
Egypt. J. Agron. (Accepted).
Sayed M.A. and R.E.E. Mahdy (2016) Heterosis and genetic parameters in grain sorghum under
irrigation and drought stress environments. Egyptian J. Plant Breed. (Accepted).
Subudhi PK, Rosenow DT, Nguyen HT (2000) Quantitative trait loci for the stay green trait in
sorghum (Sorghum bicolor L. Moench): consistency across genetic backgrounds and
environments. TheorAppl Genet101:733–741
Tag El-Din, A.A., Hessein, E.M. and Ali, E.A., 2012. Path Coefficient and Correlation Assessment
of Yield and Yield Associated Traits in Sorghum (Sorghum bicolor L.) Genotypes. AmericanEurasian Journal of Agricultural & Environmental Sciences, 12(6), pp.815-819.
Tariq M, Awan SI, Haq MIU (2007). Genetic variability and character association for harvest index
in sorghum under rainfed conditions. Int. J. Agric. Biol. 9(3): 470–472.
Tolk, J.A., Howell, T.A. and Miller, F.R., 2013. Yield component analysis of grain sorghum grown
under water stress. Field Crops Research, 145, pp.44-51.
van Oosterom EJ, Hammer GL. 2008. Determination of grain number in sorghum. Field Crops
Research 108, 259–268.
Wang, H. and Clarke, J.M., (1993) Relationship of excised-leaf water loss and stomatal frequency
in wheat. Canadian Journal of Plant Science, 73(1), pp.93-99.
15
Wanous, M.K., Miller, F.R. and Rosenow, D.T. (1991) Evaluation of visual rating scales for green
leaf retention in sorghum. Crop Sci., 31: 1691– 1694.
Warkad, Y.N., Tidke, R.T., Maske, N.M., Kharde, A.V. and Potdukhe, N.R., 2010. Character
association and path analysis in sorghum [Sorghum bicolor (L.)] Moench. International Journal
of Agricultural Sciences, 6(1), pp.100-104.
White, J. W., Ochoa, R., Ibarra, P. F., & Singh, S. P. (1994). Inheritance of seed yield, maturity
and seed weight of common bean (Phaselous vulgaris) under semi-arid rainfed conditions.
Journal of Agricultural Science, 122, 265–273.
Wright S (1921). Systems of mating. Genetics 6: 111-178
Xu, W., D.T. Rosenow and H.T. Nguyen. (2000) Stay green trait in grain sorghum: Relationship
between visual rating and leaf chlorophyll concentration. Plant Breed., 119: 365-367.
16
Table (1) Combined analysis of variance and summary statistics for grain yield, yield attributes and
physiological traits under well-watered conditions across two locations and over two seasons.
Mean squares
S.V.
d. f.
Grain yield and its attributes
GY/P
50% HD PH
PL
Physiological traits
PW
PWG
TH%
SI
193.0** 14.5
937.2
2929.3** 625.2**
Year (Y)
1
5596.9** 427.8** 595.9
Loc. (L)
1
12922** 3448**
Y*L
1
1798**
128.9** 18919** 84.3*
3.7
168.5
6865.6** 14.8*
Error a
8
150.3
4.5
3.0
238.2
59.2
Genotypes
42 1248.5** 128.9** 6432.9** 88.0**
3.3**
G*Y
42 723.5** 75.9**
1565.7** 28.2**
G*L
42 55.3
0.9
288.5** 2.1**
G*Y*L
42 41.3
1.0
Error b
336 72.13
73559** 1081.8** 48.1**
RWC
ELWL
CC
Stg
FLA
2698.5** 16594** 2353.6** 115.0** 8033.0
59.7*
200100*
4888.4** 2584.1* 110.4*
1.1
36232.0
27.4
119.1
8.3
11151.0
2737.7** 226.1** 125.9**
115.1
499.8** 45.6**
6.7**
8893.6**
1.6*
1417.3** 226.0** 56.2**
116.5
189.0** 35.4**
1.3
6620.6**
0.4
99.3
5.9
4.9
9.9
144.0** 16.9**
1.7*
904.3
188.0** 1.4
0.4
74.7
5.8
3.2
6.7
158.1** 16.5
1.3
985.4
2.8
106.4
5.3
1.0
157.9
37.1
6.1
29.6
44.5
13.3
1.0
2725.4
176.2
10.5
18302** 7444.9** 1378.4** 19235** 1412**
2.4
4161**
18.1
R2
0.81
0.93
0.92
0.78
0.50
0.79
0.74
0.83
0.79
0.80
0.72
0.66
0.53
C.V.
20.6
2.6
7.7
8.7
16.6
18.3
10.6
9.7
7.5
10.3
7.4
17.5
27.0
Mean
41.2
63.1
133.2
26.5
5.9
68.6
57.2
25.5
72.1
64.5
49.1
5.8
193.5
Minimum
21.1
54.6
92.0
19.2
4.6
38.3
47.7
18.1
66.9
49.0
44.9
3.8
138.2
Maximum
59.7
74.3
180.5
32.1
7.2
96.8
65.6
32.2
78.5
79.9
53.8
7.0
275.0
* and **; significant at P values of 0.05 and 0.01, respectively.
Table (2) Combined analysis of variance and summary statistics for grain yield, yield attributes and
physiological traits under drought stress conditions across two locations and over two seasons.
Mean squares
S.V.
d.f
Grain yield and its attributes
GY/P
50% HD PH
PL
PW
PWG
Physiological traits
TH%
SI
RWC
CC
Stg
1 43052** 848.3** 7140.7** 18.2
Loc. (L)
1 9063.1** 3354.3** 54587** 850.7** 36.5*
11233** 7872.9** 1452.6** 20302** 10670** 3821.9** 46.5*
184295**
Y*L
1 2457.1** 95.2**
15691** 75.0
3.3
604.5
7231.6** 21.1*
4047.8** 350.6
251.9** 0.4
22662*
Error a
8 140.0
191.6
1.5
203.8
57.6
123.7
10.1
3261.7
Genotypes
42 1072.8** 160.1** 3130.8** 65.7**
3.4**
1980.9** 423.2** 47.8**
136.9** 692.4** 102.6** 2.6**
7523.6**
G*Y
42 420.8** 81.4**
869.8** 43.1**
2.5**
911.5** 319.4** 29.4**
143.8** 135.0*
57.2**
0.8
7439.6**
G*L
42 65.0
0.8
137.5*
1.3
0.1
107.8
7.7
2.6
13.9
102.5
7.8
1.2
572.1
G*Y*L
42 52.4
1.2
103.0
1.4
0.1
92.9
8.4
2.5
12.9
101.6
7.0
1.0
557.1
Error b
336 68.5
2.8
86.4
7.3
0.5
153.2
43.8
6.8
39.4
87.3
12.6
1.2
3162.1
35.8
2.3
15343** 4339.7** 98.3**
FLA
Year (Y)
4.5
119.2** 44694** 19027** 264.1** 411.4
ELWL
94.0
8.2
26150*
2
R
0.84
0.93
0.89
0.70
0.73
0.78
0.81
0.69
0.74
0.71
0.78
0.52
0.46
C.V.
25.2
2.6
8.4
11.8
13.0
23.2
11.6
10.6
8.2
12.3
7.7
20.4
29.1
Mean
32.8
63.8
110.8
23.0
5.3
53.3
57.0
24.7
76.8
75.8
45.9
5.3
193.2
Minimum
13.8
57.2
74.6
17.3
4.3
30.7
40.5
21.1
65.5
49.6
41.3
4.3
139.0
Maximum
53.1
77.4
145.5
28.1
6.3
80.6
65.6
28.4
82.6
85.4
57.3
6.5
256.7
* and **; significant at P values of 0.05 and 0.01, respectively.
17
Table (3) Correlation coefficients (r) among grain yield, its attributes and physiological traits under
well-watered conditions (above) and drought stress conditions (below) over years and locations.
Trait
GY/P
50% HD
0,03
GY/P
PH PL
PW
PWG
TH%
SI
RWC
ELWL CC
0,34** 0,34** 0,48** 0,94** 0,66** 0,27** 0,36** -0,07
50% HD
0,06
-0,08
-0,19** -0,09
-0,02
0,01
-0,10*
PH
0,45** -0,22**
PL
0,32** -0,41**
0,51**
PW
0,72** -0,05
0,21** 0,31**
PWG
0,95** 0,03
0,41** 0,36** 0,70**
TH%
0,72** -0,02
0,45** 0,19** 0,55** 0,52**
SI
0,26** -0,06
0,44** 0,17** 0,21** 0,21** 0,30**
RWC
0,18** -0,30**
0,36** 0,23** 0,11*
ELWL
0,14** 0,13**
-0,12** -0,16** 0,15** 0,13** 0,07
CC
0,54** -0,14**
0,38** 0,27** 0,45** 0,48** 0,48** 0,30** 0,34** -0,04
Stg
0,24** 0,11*
0,19** 0,07
FLA
0,12** -0,12**
0,16** 0,17** 0,12** 0,16** 0,06
0,11*
Stg
0,15** 0,24**
-0,22** 0,21** -0,43** 0,00
0,64** 0,13** 0,28** 0,33** 0,51** 0,38** -0,11*
FLA
-0,19**
0,24** 0,34** 0,08
0,27** 0,36** 0,17** 0,31** 0,24** -0,21** 0,34** 0,26** 0,16**
0,48** 0,32** 0,14** 0,26** 0,01
0,17** 0,01
0,27**
0,40** 0,19** 0,24** -0,15** 0,17** 0,07
0,25**
0,30** 0,50** -0,03
0,27** 0,05
0,17** 0,21** 0,31**
-0,05
0,00
-0,26**
0,24** 0,19** 0,25** 0,25** 0,08
* and **; significant at P values of 0.05 and 0.01, respectively.
18
0,05
0,20** 0,22**
0,06
0,29** 0,10*
0,26** 0,28** 0,29**
-0,26** 0,16** -0,13**
0,03
0,26** 0,36** -0,12
-0,04
0,26**
0,17
0,23**
0,05
0,06
Table (4) Stepwise regression analysis steps of grain yield/plant and its attributes under wellwatered and drought stress conditions over two locations and two years.
Step
Source
Estimate
SE
F value
Pr > F
Intercept
-4.93
0.80
37.15
< 0.001
Panicle weight (PWG)
0.67
0.01
3597.78
< 0.001
Intercept
-30.18
0.68
1917.44
< 0.001
Panicle weight (PWG)
0.57
0.01
10331.90
< 0.001
Threshing % (TH%)
0.55
0.01
1899.99
< 0.001
Intercept
-38.03
1.52
621.61
< 0.001
Days to 50% Heading (50%DH)
0.12
0.02
32.79
< 0.001
Panicle weight (PWG)
0.57
0.01
10994.50
< 0.001
Threshing % (PWG)
0.55
0.01
2007.87
< 0.001
Intercept
-39.42
1.61
597.63
< 0.001
Days to 50% Heading (50%DH)
0.13
0.02
35.82
< 0.001
Panicle weight (PWG)
0.57
0.01
10991.70
< 0.001
Threshing % (PWG)
0.55
0.01
1847.56
< 0.001
Seed index (SI)
0.06
0.02
6.50
< 0.05
Partial R2
Model R2
87.50
87.50
9.84
97.34
0.16
97.50
0.03
97.53
91.10
91.10
6.54
97.63
0.19
97.83
Under well-watered conditions
1
2
3
4
Grain yield function equation GY/P=-39.42+0.13 50%DH+0.57 PGW+0.55 TH%+0.06 SI
Under drought conditions
1
2
3
Intercept
-6.96
0.59
138.34
< 0.001
Panicle weight
0.75
0.01
5259.02
< 0.001
Intercept
-23.36
0.53
1928.69
< 0.001
Panicle weight
0.62
0.01
10018.50
< 0.001
Threshing %
0.40
0.01
1417.65
< 0.001
Intercept
-32.08
1.38
535.64
< 0.001
Days to 50% Heading (50%DH)
0.14
0.02
45.82
< 0.001
Panicle weight
0.62
0.01
10811.80
< 0.001
Threshing %
0.41
0.01
1561.19
< 0.001
Grain yield function equation GY/P=-32.08+0.14 50%DH+0.62 PGW+0.41 TH%
19
Table (5) Stepwise regression analysis models of grain yield/plant and physiological traits under
well-watered and drought stress conditions over two locations and two years.
Step
1
2
Source
SE
F value
Pr > F
Under well-watered conditions
Estim
ate
Intercept
-2.29
5.01
0.21NS
0.64
Relative water content (RWC)
0.60
0.07
76.97
< 0.001
NS
Intercept
-4.68
4.99
0.88
Relative water content (RWC)
0.53
0.07
55.99
< 0.001
Flag leaf area (FLA)
0.04
0.01
12.85
< 0.001
Partial R2
Model R2
13.02
13.02
2.13
15.15
29.29
29.29
2.67
31.97
0.92
32.88
0.34
Grain yield function equation GY/P= -4.68+0.53 RWC+0.04 FLA
1
2
3
Under drought conditions
Intercept
-34.26
4.63
54.56
< 0.001
Chlorophyll content (CC)
1.46
0.10
212.96
< 0.001
Intercept
Excised leaf water loss (ELWL)
-5.02
0.19
5.77
0.04
75.58
20.15
< 0.001
< 0.001
Chlorophyll content (CC)
1.48
0.09
226.36
< 0.001
Intercept
Excised leaf water loss (ELWL)
-53.49
0.19
5.87
0.04
82.95
19.36
< 0.001
< 0.001
Chlorophyll content (CC)
1.41
0.10
192.68
< 0.001
Stay green (Stg)
1.31
0.49
6.98
< 0.01
Grain yield function equation GY/P=-53.49+0.19 ELWL+1.41 SPAD+ 1.31 Stg
Table (6) The final step (model) of the stepwise regression analysis of grain yield/plant, its
attributes and physiological traits under well-watered and drought stress conditions over two
locations and two years.
Partial R2
Model R2
<.0001
87,50
87,50
2071.30
<.0001
9,80
97,30
0.009
37.21
<.0001
0,30
97,60
-0.073
0.022
11.26
0.0009
0,10
97,70
Flag leaf area (FLA)
-0.005
0.002
6.74
0.0097
0,04
97,74
Days to 50% heading (50%DH)
0.058
0.023
6.30
0.0124
0,02
97,76
Seed index (SI)
0.047
0.023
4.16
0.0418
0,02
97,78
Source
Estimate
SE
F value
Pr > F
Intercept
-34.44
2.31
223.29
<.0001
Panicle weight (PWG)
0.585
0.005
12023.10
Threshing percentage (TH%)
0.553
0.012
Excised leaf water loss (ELWL)
0.055
Chlorophyll content (CC)
Under well-watered conditions
Grain yield function equation GY/P=-34,44+0.58 PWG+ 0,55 TH%+ 0,05 ELWL+ (-0,07 CC) + (-,005 FLA)
+0,05 HD + 0,047 SI
Under drought stress conditions
Intercept
-35.79
1.741
422.82
<.0001
Panicle weight (PWG)
0.614
0.006
9722.6
<.0001
91,10
91,10
Threshing percentage (TH%)
0.389
0.011
1370.8
<.0001
6,50
97,60
20
Days to 50% heading (50%DH)
0.139
0.02
48.28
<.0001
0,20
97,80
Chlorophyll content (CC)
0.099
0.021
22.3
<.0001
0,10
97,90
Flag leaf area (FLA)
-0.005
0.002
9.07
0.003
0,05
97,95
Excised leaf water loss (ELWL)
0.018
0.008
5.07
0.025
0,05
98,00
Grain yield function equation GY/P=-35,79+0.61 PWG+ 0,38 TH% + 0,13 HD + 0,09 CC+ (-0,005 FLA) + 0,018
ELWL
Table (7) Direct (bold) and indirect effects of seven yield attributes on grain yield/plant in grain
sorghum under well-watered and drought stress conditions over two locations and two years.
Independent variables
50% HD
PH
0.040**
-0.003
Pl
PW
PWG
TH%
SI
Total effect
-0.007
-0.003
-0.001
0.000
-0.004
0.022
0.000
0.000
0.000
0.001
0.003
-0.004
-0.002
-0.003
-0.026
Under well-watered conditions
Days to 50% heading
Plant height
Panicle length
0.000
0.001
0.002
ns
-0.006
0.001
-0.010
ns
-0.003
ns
Panicle width
0.001
-0.001
-0.003
-0.010
-0.005
-0.003
-0.001
-0.023
Panicle weight
-0.015
0.228
0.293
0.388
0.810**
0.324
0.157
2.187
Threshing %
0.003
0.114
0.058
0.108
0.136
0.340**
0.102
0.862
Seed index
-0.002
0.010
0.006
0.003
0.004
0.006
0.020**
0.047
0.040**
0.000
0.004
0.001
0.020
-0.007
0.001
0.058
-0.002
0.333
0.139
0.001
0.458
0.291
0.059
0.001
0.321
0.569
0.169
0.001
0.723
0.963
Under drought stress conditions
Days to 50% heading
Plant height
Panicle length
-0.009
-0.016
0.002
0.001
ns
-0.005
-0.010
ns
-0.003
ns
Panicle width
-0.002
0.001
-0.003
-0.010
Panicle weight
0.001
0.001
-0.004
-0.007
0.810**
0.162
0.001
Threshing %
-0.001
0.001
-0.002
-0.005
0.424
0.310**
0.001
Seed index
-0.003
0.001
-0.002
-0.002
0.172
0.094
0.001
0.726
ns
0.262
* and **; significant at P values of 0.05 and 0.01, respectively.
Table (8) Direct (bold) and indirect effects of five physiological traits on grain yield plant/plant in
grain sorghum under well-watered and drought stress conditions over two locations and two years.
Independent variables
RWC
ELWL
CC
Stg
FLA
Total effect
Relative water content
0.310**
0.000
0.082
0.088
0.088
0.568
Excised leaf water loss
0.000
-0.070ns
0.018
-0.011
0.009
-0.054
Chlorophyll content
-0.005
0.005
-0.020 ns
0.001
-0.005
-0.024
Stay green
0.017
0.010
-0.002
0.060 ns
0.003
0.088
Flag leaf area
0.043
-0.019
0.034
0.008
0.150**
0.216
0.030 ns
-0.044
0.172
0.008
0.014
0.180
Under well-watered conditions
Under drought stress conditions
Relative water content
21
Excised leaf water loss
-0.008
0.170**
-0.022
0.003
-0.005
0.138
Chlorophyll content
0.010
-0.007
0.510**
0.026
0.007
0.546
Stay green
0.002
0.005
0.135
0.100**
0.002
0.245
Flag leaf area
0.011
-0.020
0.086
0.006
0.040 ns
0.123
* and **; significant at P values of 0.05 and 0.01, respectively.
Table (9 Direct (bold) and indirect effects of yield attributes and physiological traits on grain yield
plant/plant in grain sorghum under well-watered and drought stress conditions over two locations and
two years.
Independent variables
HD
PH
PL
0.019*
-0.001
-0.004
-0.012
-0.003 ns -0.002
PW
PWG
TH
SI
RWC ELWL
CC
Stg
FLA
Under well-watered conditions
Days to 50% heading
Plant height
Panicle length
Panicle width
Panicle weight
Threshing percentage
Seed index
Relative water content
Excised leaf water loss
Chlorophyll content
Stay green
Flag leaf area
0.000
0.000
-0.002
-0.004
0.004
-0.008
0.000
-0.004
0.000
-0.001
-0.001
-0.001
-0.001
0.000
-0.001
-0.001
0.000
0.002
-0.032
0.111
0.007
0.003
0.001
0.002
0.002
-0.001
0.002
0.002
0.001
-0.011
0.016
0.033
-0.017* 0.575
0.281
0.137
0.169
-0.104
0.119
0.050
0.173
-0.003
0.046
0.059
0.078
0.819** 0.328
0.159
0.197
-0.121
0.139
0.059
0.202
0.001
0.019
0.010
0.018
0.022
0.336** 0.101
0.169
-0.009
0.017
0.068
0.073
0.001
0.001
0.003
0.001
0.000
0.003
0.003
0.003
0.007
-0.006
-0.006
ns
-0.024
0.130
0.080
0.037
0.050
0.077
0.011
-0.077
0.135
0.086
0.093
0.086
0.180
0.095
0.010
-0.024
0.013
0.025
-0.001
0.017
0.003
-0.006
0.000
0.043** -0.011
-0.073
0.041
0.058
0.029
0.029
0.009
0.010
0.045
-0.044
-0.027** 0.001
0.000
0.020
0.016
0.001
0.004
0.012
0.017
0.017
0.010
-0.002
0.005 ns
0.000
0.003
-0.001
-0.003
-0.005
-0.004
-0.004
-0.002
-0.005
0.002
-0.004
-0.001
-0.017*
0.001
0.001
0.001
-0.007
0.000
0.001
0.002
-0.005
0.001
0.002
-0.004
-0.002
0.132
0.002
-0.002
-0.002
0.014
0.001
-0.003
-0.006
Under drought stress conditions
0.040** 0.000
Days to 50% heading
Plant height
ns
-0.002
-0.009
-0.006ns -0.001
ns
0.003
ns
Panicle length
-0.016
-0.003
-0.003
-0.001
0.056
0.001
-0.001
-0.002
0.010
0.000
-0.003
Panicle width
-0.002
-0.001
-0.001
-0.021* -0.012
0.160
0.001
0.000
0.002
0.017
0.001
-0.002
Panicle weight
0.001
-0.002
-0.001
-0.015
0.799 ** 0.154
0.001
-0.001
0.002
0.018
0.001
-0.003
Threshing percentage
-0.001
-0.003
0.000
-0.011
0.418
0.294** 0.002
-0.001
0.001
0.018
0.002
-0.001
Seed index
-0.003
-0.002
0.000
-0.004
0.170
0.089
-0.001
-0.001
0.011
0.001
-0.005
0.005ns
ns
Relative water content
-0.012
-0.002
-0.001
-0.002
0.133
0.062
0.002
-0.004
-0.004
0.013
0.000
-0.007
Excised leaf water loss
0.005
0.001
0.000
-0.003
0.100
0.020
0.000
0.001
0.015*
-0.002
0.000
0.002
Chlorophyll content
-0.006
-0.002
-0.001
-0.009
0.382
0.141
0.001
-0.001
-0.001
0.038** 0.002
Stay green
0.004
-0.001
0.000
-0.005
0.153
0.075
0.001
0.000
0.000
0.010
0.006ns
-0.001
Flag leaf area
-0.005
-0.001
0.000
-0.003
0.130
0.016
0.001
-0.002
-0.002
0.006
0.000
-0.019**
* and **; significant at P values of 0.05 and 0.01, respectively.
22
-0.003
B
50%HD
A
-.08
-.19
-.10
PH
.64
PL
-.02 .10
.01 .28
.27
PW
-.10 .33 .36
.51 .17
.31 .34
.14
.00 .04
-.01
-.01
.55
PWG
.98
.16
.34
e
.40
TH%
.19
GY/P
.81
.02
.30
SI
Figure 1 Diagram of path coefficient analysis shows the direct, indirect and residual effects (e) of yield attributes on grain
yield per plant under A) well-watered and B) drought stress conditions.
B
A
Figure 2 Diagram of path coefficient analysis shows the direct, indirect and residual effects (e) of the physiological traits
on grain yield per plant under A) well-watered and B) drought stress conditions.
23
‫العالقة بين المحصول وكالً من مكوناته وبعض الصفات الفسيولوجية في الذرة الرفيعة للحبوب تحت الظروف‬
‫المروية والجفاف‬
‫محمد عبدالعزيز عبدالحليم سيد‬
‫قسم المحاصيل – كلية الزراعة – جامعة أسيوط‬
‫‪Email: [email protected]‬‬
‫الملخص العربي‬
‫أنه من المرغوب فيه لمربي الذرة الرفيعة للحبوب أن يعرف مدي العالقة بين المحصول والصفات المورفولوجية والفسيولوجية‬
‫أجري هذا العمل في‬
‫المؤثرة عليه والتي تسهل للمربي انتخاب نباتات ذات صفات مرغوبة خاصة تحت ظروف نقص الماء‪ .‬ولهذا‪,‬‬
‫َ‬
‫منطقتين تمثالن نوعين من التربة تابعتين للمزارع التجريبية لكلية الزراعة جامعة أسيوط بمحافظة اسيوط‪ ,‬مصر‪ ,‬خالل موسمي‬
‫‪ 2014‬و ‪ 2015‬تحت الظروف المروية والجفافية‪ .‬لتفيذ الظروف المروية‪ ,‬فانه استخدم نظام الري السطحي في التربة الطينية ورويت‬
‫النباتات كل ‪ 14‬يوم بينما استخدم نظام الري بالتنقيط في التربة الرملية حيث رويت النباتات كل ثالث أيام لمدة ساعتين‪ .‬وللحصول‬
‫علي الظروف الجفافية تم منع الرية الثالثة والخامسة في التربة الطينية بينما رويت النباتات كل ثالث أيام لمدة ساعة واحدة بنظام‬
‫الري بالتنقيط‪ .‬ثالث طرق احصائية هي معامل االرتباط البسيط‪ ,‬تحليل معامل المرور و تحليل اإلنحدار المتدرج أجريت لتحديد‬
‫العالقات بين المحصول ومكوناته وبعض الصفات الفسيولوجية تحت كالً من المعاملتين‪ 43 .‬تركيب وراثي من الذرة الرفيعة للحبوب‬
‫شملت ‪ 30‬هجينا ً وأبائها األحد عشر (‪ 6‬أمهات و‪ 5‬أباء تم التهجين فيما بينها بنظام تزاوج الساللة في الكشاف) باالضافة لصنفين‬
‫قياسيين‪ .‬أظهرت النتائج وجود اختالفات عالية المعنوية بين التراكيب الوراثية لكل الصفات المدروسة تحت كال المعاملتين‪ .‬أرتبط‬
‫وزن القنديل ارتباطا ً موجبا ً عاليا ً مع محصول الحبوب في كالً من المعاملتين تاله في قوة االرتباط نسبة التفريط ومعامل البذرة مما‬
‫يشير إلي أن هذه الصفات هي األكثر مساهمة في المحصول‪ .‬كما أشار تحليل معامل المرور أن وزن القنديل ونسبة التفريط كان لهما‬
‫تأث ير مباشر إيجابي علي محصول النبات‪ ,‬بينما كان محتوي الكلوروفيل ونسبة فقد الماء من الورقة المقطوعة و عدد األوراق‬
‫الخضراء حتي الحصاد أكثر الصفات الفسيولوجية تأثيرا ُ في صفة المحصول تحت ظروف الجفاف‪ .‬باإلضافة الي ذلك أن عرض‬
‫القنديل ومحتوي الكلوروفيل كان لهما تأثيرا ً عاليا ً غير مباشر علي محصول الحبوب للنبات تحت ظروف الجفاف‪ .‬أظهر تحليل‬
‫اإلنحدار المتدرج أن صفتي وزن القنديل ونسبة التفريط أقوي الصفات تأثيرا ً في تباين صفة المحصول تحت ظروف الجفاف‪ .‬علي‬
‫الجانب األخر‪ ,‬كان هناك ارتباط موجب معنوي بين الصفات الفسيولوجية وصفة المحصول تحت كالً من المعاملتين (ما عدا صفة‬
‫فقد الماء من الورقة المقطوعة تحت الظروف المروية)‪ .‬أشار تحليل االنحدار المتدرج أن المحتوي المائي النسبي أكثر الصفات‬
‫الفسيولوجية تاثرا ً في تباين صفة المحصول تحت الظروف المروية تلتها صفة مساحة ورقة العلم‪ ,‬بينما محتوي الكلوروفيل األكثر‬
‫تأثيرا ً في التباين الكلي لصفة المحصول تحت ظروف الجفاف تلتها صفة فقد الماء من الورقة المقطوعة‪.‬‬
‫الكلمات الدالة‪:‬‬
‫االرتباط‪ ,‬معامل المرور‪ ,‬االنحدار المتدرج‪ ,‬الجفاف‪ ,‬الذرة الرفيعة‬
‫‪24‬‬